Telescope Mirror Making

Many people are surprised to find that some club members have homemade telescopes with homemade mirrors. They are even more surprised to find out that they are better than commercial mirrors. Telescope mirror making used to be more popular 30 to 50 years ago. Commercial mirrors were not widely available and were very expensive. People had fewer distractions in life, and also less discretionary income. Also, the skies were darker and more objects could be seen.
A telescope mirror is the most precise large-sized object made by man. Surface tolerances on these mirrors must be kept to within a few millionths of an inch of perfection. It would seem logical that high-tech machines could produce a superior mirror to that which you could make at home; however, that is not generally the case. A person with nothing but two glass disks, a 55-gallon drum, abrasives, polishing pitch, and a Foucault tester can produce an essentially perfect mirror. Producing mirrors like this take time, and commercial mirror companies can only spend so much time manufacturing a mirror (after all, they have to make a buck). Over the years, manufacturers have refined their manufacturing techniques (especially with large 15 to 30-inch mirrors). Today, good commercial mirrors can be purchased from a variety of locations. But…back to the subject! Here is an outline of how to make a telescope mirror. This is a brief summary. A good mirror-making book with all the details is “How To Make A Telescope” by Jean Texereau. We will use an 8-inch f/6 as an example.
Purchase two glass disks. The mirror disk should be fine annealed Pyrex about 1.25 inches thick. The “tool” disk should be plate glass or Pyrex and be at least 0.75 inches thick. Also purchase a series of abrasives. The abrasives should be various sizes of silicon carbide and aluminum oxide. You will also need to purchase pitch, a nylon screen, a 30-watt “pencil” soldering iron, and cerium oxide polishing compound. Willmann-Bell sells mirror-making kits that have all the ingredients.
You will need to obtain an empty 55-gallon drum (preferably one that did not store plutonium, cyanide, or PCBs), a Foucault tester, (either build, buy, or borrow one), and a flashlight.
Construct a circular plywood table with feet the same diameter as the 55-gallon drum. The feet should rest solidly on the drum top surface. Set up in an area such as a utility room or basement with a utility sink nearby. There should be room to walk around the drum. Fill the drum with water after it is already in place. Place a cushion of wet newspapers on the table.
First, get a grinding stone and chamfer a 1/8-inch bevel in the mirror and tool. You want it to be big enough to not have to bevel the mirror again. Re-beveling the tool is okay. Begin grinding one glass disk against the other with the mirror on top. Use your roughest abrasive with “chordal” strokes as you slowly walk around the drum. This will hollow out the center of the mirror and begin the formation of the curve. You can use a steel ruler and a .25-inch long piece of nail that is the right diameter placed at the center of the mirror to judge the depth of the curve. If the ruler “see-saws” the curve is too shallow. If the ruler does not touch the nail, the curve is too deep.
After the required depth is reached, continue with centered 1/3 “W” strokes. This produces the spherical surface desired. Keep checking the curvature with the steel rule and nail. Mirror on top deepens the curve; tool on top makes the curve shallower.
After the surface looks completely ground, clean the area, and begin the 1/3 “W” stroke with the next finer grade of abrasive. It is very important to make sure you have all the previous abrasive cleaned up before going on to a finer abrasive. Continue on with all the finer grades of abrasive. Eight to fifteen “wets” will be required for each abrasive grade. Each “wet” starts out with a grating, grinding sound. As the grinding continues, the sound softens until no more grinding is occurring (several minutes). Gently slide the disks apart until they separate. Clean the gray paste off in a bucket of water. Then replace the tool and add more fresh abrasive. Repeat. The final aluminum oxide abrasive is very fine (like talcum powder). Be very careful not to clink the tool and mirror together even slightly. This will chip an edge. If this happens, stop immediately and clean the mirrors or the chips can gouge the glass.
After all the grinding is done, the surface of the mirror is still dull (not polished). To polish the mirror, you will have to construct a pitch lap using the tool.
To construct a pitch lap, melt the pitch and modify the softness with pine tar or turpentine. At room temperature, the pitch should be soft enough to leave a mark after a few seconds of fingernail pressure. But it should be firm enough to still seem brittle. This takes “feel” and experience. Pour the melted pitch on the tool (prepared by warming in hot water and wiping with turpentine). As it spreads out, press the mirror over it. The mirror needs to be rubbed with soapy water so the pitch won’t stick to it. Slide the mirror back and forth to mold the pitch. Cool the pitch in water. It should be about 1/4 inch thick or a little less. Heat the 30W pencil soldering iron and melt divisions in the pitch about right down to the glass tool one inch apart. The final result should look like a checkerboard with 1-inch squares. Make sure neither a square nor a channel intersection is exactly at the center of the lap. Use a sharp utility knife (with the breakable blade segments) to shave the channels open evenly. This works best with a cold lap and very sharp, new blade segment. Shave the circumference of the tool with the blade so the pitch lines up with the glass edge. Bevel the circumference edge with the blade.
Place some cerium oxide slurry on the lap surface. Place the mirror exactly on top to press the exact curve into the lap. Add some weight if desired (such as a bucket of water). When good contact is made with the lap all over, separate the disks and place the nylon window screen piece (12X12 inches is fine) over the lap. Place the mirror on top. Add the bucket for weight. Press the screen pattern into the lap (but not too deep).
You are now ready to start polishing. Use the same 1/3W strokes as grinding. After a half hour of polishing, the mirror should start to look shiny. It is far from being fully polished. An 8-inch mirror can take from 8 to 15 hours of actual polishing time. Shine a pen laser on the surface. A fully polished mirror will not show much laser image on the surface.
After the mirror is polished, the critical figuring stage begins. This is what separates the men from the boys (or the women from the girls). Until now, the process has been basically grunt work. Now, technique, knowledge, experience, and record keeping become important. Construct a mirror stand and borrow or build or buy a Foucault tester. This tester allows you to evaluate the curvature of zones in the mirror using a series of masks. The correct surface is parabolic, and different radial zones in the mirror will “null” at particular knife-edge readings. The trick is to be able to test the mirror, and apply various corrective strokes to it using the lap. When fully corrected, the mirror image in the Foucault tester should show a smoothly corrected paraboloid, and the knife-edge readings at various radial zones should be very close to the calculated ideal. There are computer programs that will calculate the wavefront error using inputted knife-edge readings. You will become very familiar with optical errors such as “the dreaded turned down edge”, “dog biscuit”, “zones”, “astigmatism”, “under and over correction”, “spherical aberration”, and many others. Strive to make a smooth mirror with better than a 1/16th wave error on the reflected wavefront (1/32 wave surface error). This can be time consuming and frustrating, but it is possible. Details for testing are presented by Texereau very well.
If you can succeed at this, you will do so with the knowledge that very few commercial mirrors are this good. You will have created one of the most precise large surfaces in the world with nothing but your own two hands. Now, the mirror must be packed carefully and sent to be aluminized.